Diagnostic Performance of US for Differentiating Perforated from Nonperforated Pediatric Appendicitis: A Prospective Cohort Study.

Abstract

Purpose To prospectively evaluate the diagnostic performance of ultrasonography (US) for differentiating perforated from nonperforated pediatric appendicitis and to investigate the association between specific US findings and perforation. Materials and Methods This HIPAA-compliant study had institutional review board approval, and the need for informed consent was waived. All abdominal US studies performed for suspected pediatric appendicitis at one institution from July 1, 2013, to July 9, 2014, were examined prospectively. US studies were reported by using a risk-stratified scoring system (where a score of 1 indicated a normal appendix; a score of 2, an incompletely visualized normal appendix; a score of 3, a nonvisualized appendix; a score of 4, equivocal; a score of 5a, nonperforated appendicitis; and a score of 5b, perforated appendicitis). The diagnostic performance of US studies designated 5a and 5b was calculated. The following US findings were correlated with perforation at multivariate analysis: maximum appendiceal diameter, wall thickness, loss of mural stratification, hyperemia, periappendiceal fat inflammation, periappendiceal fluid, lumen contents, and appendicolith presence. The number of symptomatic days prior to presentation was recorded. Surgical diagnosis and clinical follow-up served as reference standards. Results A total of 577 patients with a diagnosis of appendicitis at US met the study criteria (468 with a score of 5a; 109 with a score of 5b). Appendicitis was correctly identified in 573 (99.3%) of 577 patients. US performance in the detection of perforated appendicitis (5b) was as follows: a sensitivity of 44.0% (80 of 182), a specificity of 93.1% (364 of 391), a positive predictive value of 74.8% (80 of 107), and a negative predictive value of 78.1% (364 of 466). Statistically significant associations with perforated appendicitis were longer duration of symptoms (odds ratio [OR] = 1.46, P < .0001), increased maximum diameter (OR = 1.29, P < .0001), simple periappendiceal fluid (OR = 2.08, P = .002), complex periappendiceal fluid (OR = 18.5, P < .0001), fluid-filled lumen (OR = 0.34, P = .002), and appendicolith (OR = 1.67, P = .02). Conclusion US is highly specific but nonsensitive for perforated pediatric appendicitis. Several US findings are significantly associated with perforation, especially the presence of complex periappendiceal fluid. © RSNA, 2016.